OCTOBER
1956
1179
XOTES
procedure11 for the preparation of a 3,5-dinitrobenzoate1 0.15 g. of a cream-colored solid was obtained, which melted at 10'3-111" :tnd did not depress the m.p. of an authentic sample of i,enzyl 3,j-dinitrobenzoate [reported m.p. 12 113-113.5" icorr. 11. DEPBRTMEST O F CHEMISTRY STATEUSIVERSITEO F Iowvl Iowa CJTY.IOWA - ____(11) J. H. Brewster and C. J. Ciotti, Jr., J . Am. Chem. Soc., 77, 6214 (1955). (12) R, E. Buckles and J. E, I\laurer, J . Org, Chem., 18, 1585 (1953).
Occurs despite the very extensive hydrogen bonding, presumably inl,olving the carbonyl grOUp, that is present in these compounds* Of the examined, only p-phenylacetanilide exhibits an . _ . unusual carbonyl absorption, which may arise from partial hydrogen bonding. TABLE I
CARBONYL STRETCHIXG FREQUENCIES OF ACETAXILIDES A,
Benzene Ring D. G. O'SCLLIVAN ASD P. W. SADLER
-0.660 -0.268 -0.170 -0 069 0
p-Ph
0,009 0.062 0.115 0.227 0.337 0.352 0.373 0.710
m-Me0
The carbonyl group possesses a stretching frequency in the 1800-1650 cm.-l region of the infrared spectrum.',* Certain sets of compounds containing the substituted benzoyl group give carbonyl maxima which are linear functions of the u values of the s u b ~ t i t u e n t s . ~The ~ ~ J latter numbers are monotonically related to the decrease of electron density in the region of the particular benzene-ring carbon atom, produced by a substituent in either the meta- or para-position.6 The effect of substituents may, however, be further transmitted through an n" group attached to the benzene ring as is shown by the tabulated carbonyl frequencies of meta- and para-substituted acetanilides and isonitrosoacetariilides. For comparison, carbonyl frequencies of substituted acetophenones and benzop h e n o n e ~are ~ also included in Table I. Amongst other compounds that shorn this effect are benzoic acid monomers and dimers, benzoyl peroxides,' and isatins.8 The carbonyl frequencies of all these compounds in solution approximately obey the equation v -- Y O = 15u, where vo = 1684 (acetanilides), 1680 (isonitrosoacetanilides), 1690 (acetophenones) , 1663 (benzophenones), 1742 (benzoic acid monomers), 1696 (benzoic acid dimers), and 17-10 (isatins). The good correlation between carbonyl frequency and JS function in acetanilides (1) Bellamy, The Infra-red Spectra of Complex Molecules, Methuen and Co., London, 1954. (2) Jones and Sandorfy, Chemical Applications of Spectrosco~v(Editor: W. West ), Interscience Publishers, New Yo& isis, p. 443. (31 Flett. Trans. Faradav SOC..44. 767 (1948). i4i Fuson. Josien, and ghelton, J. Am. Chem. SOC.,76, 2526 (1954): ( 5 ) Hammeti,. Phvsical Oraanic Chemistrv. McGraw-Hill Bdok Co., New Y o & 1940, p: 188. (6) Jaffe, J . Chem. Phys., 2 0 , 279 (1952). (7) Davison, J . Chem. Soc., 2456 (1951). (8) O'Sullivan and Sadler, J . Chem. SOC.(in the press). " I
p-c1 m-F m-I
m-C1 m-NOz a
A"
U
p-r\"z p-Me0 p-Me m-Me ?;one
P-F
Received May d l , 1966
c~
ACEToPHENoNES
BENZOPHENONES D, WITH THE u VALUESOF THEIR SUBSTITUENTS
Substituent
Transmission through the PhenylaminoGroup of the Effect of Substituents in the
B,
lSOXITRoSOACETASILIDES
1678 1682 1683 1685 {::ish. 1685 1686 1688 1689 1690 1690 1696
Carbonyl Frequencies B" C* Dh -___ 1677 1651 1684 1658 1677 1661 1679 1680 1692 1664 16&0 1682 1683 1683 1685 1686 1690
1692
1666
1701
In chloroform. In carbon tetrachloride.
Certain physical properties and many chemical reaction rates may be correlated with u values of substituents. As an illustration, in Table 11, rates of de-acylation of substituted o-nitroacetanilidesg-12 are compared with the carbonyl frequencies of the corresponding acetanilides. It has been shown that the nitro-group in the ortho-position exerts a constant effect in all cases, so that the comparison is ~alid.~-'~ TABLE I1 U-VALUES AND CARBOXYL FREQUENCIES ( Y ) OF ACETANILIDES COMP.4RED WITH RATESO F DE-ACYLATION (0) O F CORRESPONDING 0-NITROACETANILIDES Substituent P-NH? p-Me0 P-AIP~N p-Me m-Me None p-c1 p-Br m-C1 m-Br m-NG p-NOz
P
U
-0 -0 -0 -0 -0 0 0 0 0 0 0 1
660 268 205 170 069 227 232 373 391 710 27"
0 0 0 0
V
06 6 06 6 1 1 4
4 10 10 12 18
1678 1682 1683 1685 1688 1690 1696
Value quoted by Hammetts for derivatives of p-nitroaniline. (9) Wepster and Verkade, Rec. trav. chim., 67, 411 (1948). (10) Wepster and Verkade, Rec. trau. chim., 67, 425 (1948). (11) Wepster and Verkade, Rec. trav. chim., 68,77 (1949). (12) Wepster and Verkade, Rec. trav. chim., 68,88 (1949).
1180
NOTES
The results obtained by Richards and Thompon some substituted acetanilides in dioxane, suggest that similar carbonyl frequency shifts occur in this solvent. However, the work of these authors13 shows that much bigger shifts occur with solid acetanilides and this is also apparent with solid acetophenones. Substituted ox in dole^'^ have carbonyl maxima that obey the equation v - vo = 37.5,. v o possessing the value 1725 cm.-' It is likely that the augmentation of the effect in the latter cases arises from the absence of free rotation about the Ph--S--C bonds in oxindoles, and in acetanilides and acetophenones in the solid state. As a result, the position of the carbonyl group is fixed, presumably in the plane of the benzene ring. The (loplanarity assists the transmission of substitueqt effects. It is not clear why the shifts are small with substituted isatins. Certain other features, present in the infrared spectra of substituted acetanilides in chloroform, now are briefly mentioned. All these compounds exhibit a single, very broad, band in the 35003200 cm. region, indicating that hydrogen bonding (presumably intermolecular) is very extensive in thir s o l ~ e n t . 'In ~ most cases there was no evidence of any distinct maximum arising from an unassociated NH group, and, consequently, it was not possible to gain any information on the effect of substituents on the N H frequency. It is of interest to note, however, that Flett3 has examined the KH2 frequencies of substituted anilines in carbon tetrachloride. The symmetric and asymmetric stretching vibrations of the KH2 group give rise to two maxima in the 3535-3400 cm.-l region. Substituents have a marked influence on the frequency maxima, the effect being more pronounced on the asymmetric mode. In each case, an approximately linear relationship exists between NH2 frequency and u value for those substituents possessing a value greater than -0.2. The aromatic C-C stretching vibration is present in acetanilides close to 1610 cm.-l, and this maximum undergoes splitting when meta- or weakly ortho-para-directing substituents are attached to the ring. Unassigned bands appear in the 13501300 ern.-' region, some of which may be related to the Ph--N stretching mode.
,
EXPERIMENTAL
The infrared spectra were determined in chloroform using a Perkin-Elmer 21 double-beam recording spectrometer with a rock-salt prism.
COURTACLD INSTITUTE OF BIOCHEMISTRY, THEJ ~ D D L EHOSPITAL, SEX LONDON, W.1, ENGLASD. -___
(13) Richards and Thompson, J . Chem. SOC.,1259 (1947). (14) Soloway and Friess, J . Am. Chem. Soc., 73, 5000
(1961). (15) Kellie, O'Sullivan, and Sadler, J . Chem, Soc. (in the
press).
VOL.
21
The Nuclear Magnetic Resonance Spectrum of Helvolic Acid NORMAX L. ALLINGER~ Received May 63, 1956
A tentative structure was recently proposed for helvolic acid2 in which all of the known groupings mere accounted for with the exception of the location of one carbon atom. It is the purpose of this report to offer evidence for the position of this remaining atom. The C-methyl analysis of helvolic acid gave a value of 3.62 C-methyls per mole.2 This number is rather high to represent four, and can best be taken as indicative of five such methyls per molecule. The formula proposed earlier2 TT-ould be predicted to give four C - m e t h y l ~ ,two ~ from the acetoxyl groups, one from C18, and one from the acetone resulting from cleavage of the side chain. The extra carbon atom would thus appear likely to be present as a methyl group. Ordinarily extra carbon atoms in steroids are found as alkyl groups at CZ4.The ultraviolet spectra of helvolic acid and its derivatives did not, however, appear consistent with a methyl group a t this position. On the other hand, there was available some evidence which appeared to indicate that no hydrogen atom was present at (224. There was no hydrogen out-of-the-plane bending band near 12 p in the infrared spectrum, and there was no strong band in the 6p region which could be attributed to the carbon-carbon double bond stretching absorption. The stretching absorption is ordinarily very intense for a double bond of the mesityl oxide type,* while it might be expected to be weak if a nearly symmetrical arrangement around the double bond existed, as it would if a Cz4methyl were present. In view of the conflicting evidence, location of the extra carbon was not specified in the original formulation. Recently nuclear magnetic resonance spectra have been used to locate the positions of double bonds in molecules by determining the number of hydrogens present on the double bonds,3 and the success of this technique is dependent on the fact (1) National Science Foundation Post Doctoral Fellow, 1955-1956. Present address: Chemistry Department, Wayne State University, Detroit, Michigan. (2) Cram and rlllinger, J . Am. Chem. Soc., 78, in press (1956). (3) Eisenbraun, McElvain, and Agcork, J . Am. Chem. SOC.,7 6 , 607 (1954). (4) (a) Bellamy, The Infrared-red Spectra of Complex Molecules, John Wiley and Sons, Inc., Xew York, 1954, p. 37; (b) Djerassi, Bowers, and Khastgir, J . A m . Chem. SOC., 78, 1729 (1956). (5) (a) Dauben and Hance, J . Am. Chem. SOC.,77, 2451 (1955); (b) Corey, Burke, and Remers, J . Am. Chem. Soc., 77, 4941 (1955); (c) Ettlinger and Kennedy, Chemistry & Industry, 166 (1956).
